Laboratory Building Energy Analysis

Transcription

1 Laboratory Building Energy Analysis Course No: M4-11 Credit: 4 PDH Steven Liescheidt, P.E., CCS, CCPR Continuing Education and Development, Inc. 9 Greyridge Farm Court Stony Point, NY 198 P: (877) F: (877) info@cedengineering.com

2 L ABORATORIES FOR THE 21ST C ENTURY: ENERGY ANALYSIS U.S. Department of Energy Energy Efficiency and Renewable Energy

3 Laboratories for the 21 st Century: Energy Analysis Prepared for Laboratories for the 21 st Century a joint program of the U.S. Environmental Protection Agency and the U.S. Department of Energy Office of Energy Efficiency and Renewable Energy Federal Energy Management Program Prepared By Enermodal Engineering, Inc. And the National Renewable Energy Laboratory A DOE national laboratory DOE/GO April 23

4 INTRODUCTION Because laboratories use a high amount of energy often more than five times as much per square foot as an office building it is important to find cost-effective ways to reduce their energy use and costs. This energy analysis was performed to evaluate selected energy efficiency measures for a generic laboratory building. Using a computer model, the analysis compared results for a base case laboratory with results for laboratories in four different climates those of Atlanta, Denver, Minneapolis, and Seattle. The analysis focused on efficiency strategies designed to reduce the considerable amount of energy used in ventilating, cooling, and heating laboratory buildings. The impacts of humidity controls and plug load assumptions on energy loads were also considered. Results are presented and discussed in this report. Enermodal Engineering, Inc., performed the analysis, along with staff in the U.S. Department of Energy (DOE) National Renewable Energy Laboratory (NREL). This study was conducted in support of Laboratories for the 21 st Century, a joint program of the U.S. Environmental Protection Agency (EPA) and DOE through the DOE Federal Energy Management Program in the Office of Energy Efficiency and Renewable Energy. Labs 21 encourages the design, construction, and operation of safe, sustainable, high-performance laboratories. ACKNOWLEDGMENTS Sue Reilly of Enermodal Engineering and Otto Van Geet of NREL made significant contributions to the analysis and report. They would like to thank Phil Wirdzek of EPA, Will Lintner of DOE, and Nancy Carlisle of NREL for support and thoughtful reviews. They also are grateful to Paul Matthew of Lawrence Berkeley National Laboratory for his detailed review and helpful suggestions. 2

5 EXECUTIVE SUMMARY This study analyzes the effects of energy efficiency measures in a simplified laboratory model in the climates of Minneapolis, Denver, Seattle and Atlanta. The analysis establishes a base case to which certain energy efficiency strategies are applied, as well as changes in humidity controls and plug load assumptions. The analysis compares energy use and costs, chilled water and hot water system sizing, and lifecycle costs. The laboratory model is a four-story, 1,-square-foot (sf) building with 7% of its area devoted to laboratories. The window-to-wall ratio is.25 and the windows are distributed equally around the building. The insulation levels and window energy performance values meet the ASHRAE building energy standard, as does the lighting power density. The equipment power density for the building is 9 watts per square foot (W/sf) under design conditions. The cooling setpoint is 74 F and the heating setpoint is 72 F. A constant-volume reheat system serves the building with a maximum relative humidity of 6% and a minimum relative humidity of 3%. Outside air ventilation is supplied at a minimum of 2 cubic feet per minute per square foot (cfm/sf) by premium-efficiency fans. The building has a central plant with water-cooled chillers and hot water boilers. The chillers are rated at.5 kilowatt (kw)/ton and the boilers are 8% efficient. All pumps are high efficiency and have variable-speed drives. Measured and predicted energy use data from laboratory case studies were used to tune the simulation models. The simulation models are within 5% of the electricity use measured in the Labs 21 case studies (21). However, gas usage in the simulations is comparatively high, and the case studies do not provide sufficient detail to explain the differences. We estimated the space heating load from the outside air ventilation load and found good agreement with simulation results when we excluded humidity controls. We concluded that the difference in energy use is attributable to humidity controls, weather, internal loads, operating hours, and the high density of laboratory space. The table below shows the simulation results for building energy use. Seattle has the mildest climate and the lowest annual energy cost; Atlanta has the lowest total energy use. Electricity rates of $.3 per kilowatt-hour (kwh), $7/kW on-peak, and $4/kW offpeak were used. On-peak hours are 8 a.m. to 1 p.m., Monday through Friday. Gas rates are $.6/therm. Peak Demand (W/sf) Building Energy Use Annual Electricity Cost ($/sf) Annual Gas Cost ($/sf) Total Energy Cost ($/sf) Electricity Gas Total (kwh/sf/yr) (kbtu/sf/yr) (kbtu/sf/yr) Minneapolis $4.3 $5.2 $9.5 Denver $4. $4.8 $8.8 Seattle $3.9 $2.6 $6.5 Atlanta $4.7 $2.2 $6.9 3

6 Energy efficiency strategies included reducing the air flow during unoccupied periods, a variable-air-volume () system; lower static pressure drop in the air distribution system; energy recovery by enthalpy wheels, heat pipes, and run-around loops; evaporative cooling; and more accurate accounting for plug loads. Other strategies, such as reducing lighting loads and solar heat gain, were not addressed. Furthermore, we did not quantify the impact of high-efficiency equipment such as chillers, boilers, fans, pumps, and motors. Results show that the most efficient measures are the same for all climates with the exception of Denver s, where evaporative cooling is also cost-effective. Predicted energy savings differ from climate-to-climate. d on the simulation results, we conclude the following: Using a system (e.g., fume hoods) rather than a constant-volume system has the potential to reduce fan energy and energy for space cooling and heating. Energy cost savings average $1/sf in all four climates. Some form of energy recovery should always be considered. Because of the sensible and latent energy recovery achieved with an enthalpy wheel, it is the most efficient of the energy recovery alternatives considered here. The increase in fan energy from energy recovery ventilation systems is not offset by the reduction in space cooling. However, the lower heating energy use more than compensates for the increase in fan energy. Energy recovery can potentially reduce the size of the heating and cooling equipment, and a system has the potential to reduce the size of the heating system. The first-cost savings can cover a large portion of the cost of the energy efficiency strategy. Because of the high ventilation requirements in laboratory buildings, the air distribution system should be optimized to minimize pressure drop through the system and reduce energy use. Humidity control is energy-intensive and should be carefully integrated into the control strategies to minimize reheat and subcooling. Plug loads and internal gains from plug loads should be accurately assessed in order to design the mechanical system and determine power requirements. Significant increases in first costs and operating costs result from assuming too high a design load. 4

7 TABLE OF CONTENTS Chapter 1. Building Energy Simulation Case 1.1 Energy Simulation Program 1.2 Climates Selected 1.3 Utility Rates 1.4 Simulation Model 1.5 DOE-2.2 Simulation Results Chapter 2. Energy Efficiency Strategies 2.1 Ventilation Rates 2.2 Fan Static Pressure 2.3 Energy Recovery 2.4 Evaporative Cooling 2.5 Humidity Controls 2.6 Plug Loads 2.7 Summary of the Simulation Results Chapter 3. Cost-Effectiveness Analysis 63 Chapter 4. Conclusions 4.1 Peak Electricity Demand 4.2 Electricity and Gas Use 4.3 Energy Costs 4.4 Downsizing HVAC Equipment 4.5 Economics References 8 5

8 CHAPTER 1. BUILDING ENERGY SIMULATION BASE CASE Staff in the National Renewable Energy Laboratory Federal Energy Management Program (FEMP) contracted with Enermodal Engineering, Inc., to analyze energy efficiency measures for the Laboratories for the 21 st Century ( Labs 21 ) program. The purpose of the study was to identify cost-effective energy efficiency measures for a generic laboratory in different climates. Task 1 is presented in this chapter; it describes the base case simulation model and building energy simulation results. 1.1 Energy Simulation Program The building energy analysis was performed using the DOE-2.2 building energy simulation program. DOE-2.2 is an hourly simulation program that was developed by James J. Hirsch and Associates and Lawrence Berkeley National Laboratory (PC DOE h, 22). Version 41 of DOE-2.2 includes models for energy recovery ventilation, improved chiller part-load operation, and water-side economizers. 1.2 Climates Selected Four climates associated with four metropolitan areas were selected: Minneapolis, Denver, Seattle, and Atlanta. Seattle has the mildest climate conditions, Minneapolis the coldest, and Atlanta the warmest. Denver has a dry climate; Atlanta and Minneapolis are humid in summer. Table 1.1 shows winter and summer design temperatures and heating degree days (HDD) and cooling degree days (CDD) from the weather tape. Table 1.1 Design Conditions Design Conditions Winter design temperature ( F) Summer design temperature ( F) 88/77 9/59 81/64 91/74 HDD65 (F-days) CDD65 (F-days) For the simulations, TMY2 hourly weather data are used for all climates except Denver s. No TMY2 data were available for Denver, so TMY ( typical meteorological year ) data were used. Altitude is not included in the TMY tape for Denver, so it was input to the DOE-2 file. 1.3 Utility Rates Laboratory buildings are generally high-demand buildings with greater than 1 W/sf of peak demand. Electricity rate structures vary around the country for such high-demand buildings (>5 kw peak demand). Typically, there is an energy charge ($/kwh) and a 6

9 peak demand charge ($/kw), and charge rates may vary with the time of day, time of year, amount of energy used, or all three. For this study, we assumed a constant energy charge of $.3/kWh, plus an on-peak demand charge of $7/kW and an off-peak demand charge of $4/kW. On-peak hours are 8 a.m. to 1 p.m., Monday through Friday. A fixed monthly charge of $15 was also included. For natural gas, a rate of $.6/therm was assumed with a fixed monthly charge of $ Simulation Model Shell Figure 1.1a EQUEST rendering of building. The model laboratory building has four stories above grade (Figure 1.1a). It has a total of 1, square feet (sf) with 25, sf on each floor. The floor-to-ceiling height is 9 feet and the floor-to-floor height is 15 feet. The building has a window-to-wall ratio of.25, and windows are distributed equally around the building. Table 1.2 provides details on the building shell, such as wall and window areas, and insulation levels. The insulation levels and window performance values are based on ASHRAE prescriptive requirements for Minneapolis and Denver. The only difference in the 7

10 standard requirements for Seattle and Atlanta is R-13 insulation in the walls. To simplify the simulation model, the higher wall insulation level was used for all locations. Table 1.2 Building Details Building Component Model Assumption Floor Area (sf) 1, Number of Stories 4 Floor-to-floor Height (ft) 15 Floor-to-ceiling Height (ft) 9 Net Wall Area (sf) North 9487 East 9487 South 9487 West 9487 Window Area (sf) North 593 East 593 South 593 West 593 Window-to-Wall Ratio.25 Window Shading None Wall Construction Insulation R-13+R-3.8 c.i. Total U-Value.84 Roof Construction Built-up Insulation R-15 c.i. Total U-Value.63 Slab 8 Concrete Total R-Value No Insulation Window U-Factor (Btu/hr-ft 2 -F).57 Window Solar Heat Gain Coefficient.39 (all).49 (north) Internal Loads and Lighting Rather than assume that 1% of the area in the building is laboratory space, we included ancillary spaces to make the model more realistic. The building space is thus divided into 7% laboratory area, 2% corridor, 5% restrooms, and 5% mechanical and electrical (ME) rooms. It is occupied between 8 a.m. and 1 p.m., and occupancy varies with time of day, as do the equipment and lighting schedules (Figures 1.2a, 1.2b, and 1.2c). The equipment power density (i.e., plug loads) and lighting power density vary with Figure 1.1b Perimeter and core zones in the building and their associated equipment power densities (EPD) and lighting power densities (LPD). 8

11 space type (Table 1.3). We assumed that 81% of the perimeter is laboratory space and 19% consists of support spaces. In the core, we assumed that 67% is laboratory space and 33% is support spaces. The average equipment power density for the support spaces is 1.25 W/sf. Performing an area-weighted calculation results in 1 W/sf of equipment plug loads in the perimeter zones and 8.5 W/sf in core zones (Figure 1.1b). The area-weighted average equipment load for the building is 8.8 W/sf; adjusted for the equipment schedule, it is 7 W/sf. Table 1.3 Internal Loads % Area in Building Internal Load Model Assumption Equipment Power Density (W/sf) Laboratories 7% 12 W/sf Corridor and Lobby 2% 1.25 W/sf Restrooms 5%.5 W/sf ME Rooms 5% 2 W/sf Equipment Schedule 1-8: 5%, 9-17: 8%, 18-24: 5% Number of Occupants (sf/per) 275 Occupancy Schedule 1-8: 5%, 9-1: 2%, 11-12: 95%, 13-14: 5%, 15-18: 95%, 19: 3%, 2-22: 2%, 23-24: 5% Lighting Power Density (W/sf) Laboratories Corridor and Lobby Restrooms ME Rooms 7% 1.8 W/sf 2%.7 W/sf 5% 1. W/sf 5% 1. W/sf Lighting Schedule 1-7: 1%, 8: 5%, 9-18: 9%, 19-22: 5%, 23-24: 1% Ballast Type Electronic Ballast Power Factor.9 The lighting power density assumptions are based on ASHRAE prescriptive requirements. The perimeter zones have a lighting power density of 1.8 W/sf, and the core zones are 1.44 W/sf (Figure 1.1b). The lighting power density for the building under design conditions is 1.5 W/sf; adjusted for the lighting schedule, it is 1.4 W/sf. Percent Equipment Use 9% 8% 7% 6% 5% 4% 3% 2% 1% % Equipment Schedule Percent Occupancy 1% 9% 8% 7% 6% 5% 4% 3% 2% 1% % OccupancySchedule Figure 1.2a and b Equipment and occupancy schedules. The time represents the hour before the number (e.g., 8 a.m. represents 7 a.m. to 8 a.m.). 9

12 Percent Lighting Use 1% 9% 8% 7% 6% 5% 4% 3% 2% 1% % Lighting Schedule Figure 1.2c Lighting schedule. The time represents the hour before the number Mechanical System For the mechanical system, the assumptions (see Table 1.4) are nearly identical for each climate except for the sizes of the heating, ventilation, and air-conditioning (HVAC) equipment. The HVAC system is a constant-volume reheat system. The building has four perimeter zones and a core zone on each floor (Figure 1.1b). The room temperature is maintained at 72 F for heating and at 74 F for cooling. The relative humidity is controlled at a minimum of 3% and a maximum of 6%. Supply air is 1% outside air at a minimum of 2 cfm/sf in all spaces, although in an actual building the support spaces would be handled separately. The supply air temperature is reset based on the outside air temperature and the zone calling for the maximum heating or cooling. The design air temperature leaving the main cooling and heating coils is 55 F. As the outside air temperature drops below 8 F, the supply air temperature is adjusted up and is set as high as 65 F when the outside air is 6 F or lower. There are three supply air fans and a manifold exhaust system. The perimeter zones are served by one fan, as are the first- and second-floor core zones and the third- and fourth-floor core zones. The minimum design air flow of 2 cfm/sf is based on having a fume hood every 45 sf. The fume hoods are assumed to have an average sash height of 18 inches and are 6 feet wide. A face velocity of 1 feet per minute (fpm) is maintained. The average exhaust is 9 cfm, which translates into an exhaust rate of 2 cfm/sf. We allowed DOE-2 to size the fans and found that the internal gains are the determining factor in calculating the design flow rates, not the exhaust requirements. The perimeter zones have higher internal gains than the core zones; therefore, the design flow rates are higher in the perimeter zones (Table 1.5). Note that Denver s high altitude results in higher flow rates than those of the other climates. The fans are specified to meet ASHRAE requirements of.8 W/cfm for supply fans, and they are assumed to be vane axial with premium motors. The supply fans experience 5.2 inches (water gauge) (in. w.g.) of total static pressure and have a fanplus-motor efficiency of 76% (the fan is 8% efficient and the motor is 95% efficient). 1

13 The total static pressure in laboratories is often 6 in. w.g. or more, which makes it difficult to meet ASHRAE fan requirements. The exhaust fans experience 2 in. w.g. of total static pressure and also have a total efficiency of 76%. The exhaust fans use.3 W/cfm. The chillers are high-efficiency (.5 kw/ton), centrifugal water-cooled chillers. The DOE-2.2 part-load efficiency curve was used to predict performance. High-efficiency chillers should be specified to optimize performance within the range in which the chiller most frequently operates. The chilled water is supplied at 44 F, and outside air reset controls are assumed. The supply temperature of the chilled water is set up as the outside air temperature drops below 8 F. The chilled water temperature is set at 55 F when the outside air is 6 F or below. The primary and secondary pumps have variable-frequency drives (VFDs) and high-efficiency motors. The cooling tower has two cells and two-speed fans and is sized by DOE-2. Standard gas-fired boilers with an efficiency of 8% are assumed for water heating. Although high-efficiency boilers are not included as one of the efficiency measures, they are worth considering. All pumps have VFDs and high-efficiency motors. Table 1.4 Mechanical System Mechanical System Model Assumptions System Constant-volume w/ hot water reheat Supply Air Handling Units 3 CV units; 2 cfm/sf minimum, 76% fan+motor efficiency Exhaust Air Handling Units 3 CV units (manifold exhaust), 76% fan+motor efficiency Supply Static Pressure 5.2 in. w.g. Exhaust Static Pressure 2 in. w.g. OA Ventilation Rate 1% (minimum 2 cfm/sf) Humidification 3% winter; 6% summer Cooling Thermostat Setpoint 74 F Chillers Centrifugal chillers (water cooled) Chilled Water Supply Temperature 44 F Chiller Tonnage and Number 2 (2) Chiller kw/ton.5 kw/ton Chiller Part Load Performance DOE-2.2 Default Primary Chilled Water Pumps VFD (2) Secondary Chilled Water Pumps VFD (2) Cooling Towers 1 open, 1 cell, 2 speed Condenser Water Pumps VFD (2) Economizer Cycle Air-side Heating Thermostat Setpoint 72 F Heating System Hot water boilers (2) supplying main air handlers Hot Water Supply Temperature 135 F Supply Air Temperature Leaving Main Hot Water Coil 55 F Boiler Efficiency 8% Primary Hot Water Pumps VFD (2) Secondary Hot Water Pumps VFD (2) 11

14 The sizes of the chillers, pumps, cooling tower, and boilers depend on the climate. Design conditions were assumed with 1% equipment loads, lighting use and occupancy. DOE-2.2 design-day calculations were included to determine the size of the HVAC equipment. However, DOE-2.2 does not adequately size the chillers and boilers, because peak conditions occur in Minneapolis and Atlanta when there is a significant latent load at off-design load conditions (i.e., lower dry bulb and higher wet bulb temperatures). Multiple runs were performed to determine the chiller and boiler sizes that minimized the number of hours in which the zones are underheated and undercooled to fewer than 2 hours. The design cooling and heating capacities for each zone were then fixed, and DOE-2 sized the loops, associated pumps, and cooling tower. Table 1.5 lists the sizes of the equipment used for each climate. Table 1.5 HVAC Equipment Sizes Equipment Supply and Exhaust Fans Core: 2 cfm/sf Perimeter: 2.7 cfm/sf Core: 2.4 cfm/sf Perimeter: 3.2 cfm/sf Core: 2 cfm/sf Perimeter: 2.6 cfm/sf Core: 2 cfm/sf Perimeter: 2.7 cfm/sf Chillers 65 tons (2) 4 tons (2) 35 tons (2) 65 tons (2) Chiller Primary Pumps 156 gpm (a) 96 gpm (2), 84 gpm (2), 156 gpm (2), Chiller Secondary Pumps (2) Condenser Water Pumps (2) (2), 24 ft 26 gpm (total), 5 ft 356 gpm (total), 74 ft 24 ft 16 gpm (total), 5 ft 219 gpm (total), 74 ft 2Ft 135 gpm (total), 5 ft 192 gpm (total), 74 ft 24 ft 2615 gpm (total), 5 ft 356 gpm (total), 74 ft Cooling Tower (2 cells 15 tons 92 tons 8 tons 15 tons w/ 2-speed fans) Boilers 45 hp (2) 4 hp (2) 25 hp (2) 25 hp (2) Hot Water Primary Pumps 15 gpm (2), 6 ft 134 gpm, 6 ft (2) 84 gpm (2), 6 ft 84 gpm (2), 6 ft Hot Water Secondary Pumps (2) 235 gpm (total), 38 ft 214 gpm (total), 38 ft (a) gpm = gallons per minute. 358 gpm (total), 38 ft 32 gpm (total), 38 ft 1.5 DOE-2.2 SIMULATION RESULTS Tuning the Models The laboratory building was simulated in each of the climates using the DOE-2.2 building energy simulation program. Table 1.6 gives the peak electricity demand and energy intensities for the building in each climate. For reference, office buildings typically operate at less than 1 W/sf peak demand and use less than 1, British thermal units (1 kbtu)/sf. Laboratories have been known to consume 5-1 times that much energy. 12

15 Table 1.6 Energy Intensities Peak Demand Electricity Gas Total (W/sf) (kwh/sf/yr) (kbtu/sf/yr) (kbtu/sf/yr) Minneapolis Denver Seattle Atlanta FEMP has prepared a number of case studies on laboratories around the country. Table 1.7 presents measured data from two of these case studies for comparison with the model developed for this analysis. Both laboratories in the case studies employed significant energy efficiency measures. The predicted energy savings for the Process and Environmental Technology Laboratory are 4%, and the electricity energy savings for the Fred Hutchinson laboratory are 33%. Table 1.7 Energy Use in Case Studies Case Study Annual Annual Gas Use Total Site Notes Electricity Use (kbtu/sf/yr) Energy (kwh/sf/yr) (kbtu/sf/yr) Process and (428 before Original study Environmental efficiency predicted a total of Technology Laboratory, measures) 595 kbtu/sf/yr Albuquerque, NM Fred Hutchinson 49 (73 before Limited information Cancer Research efficiency on gas use Center, Seattle, WA measures) We compared the results of the simulation model with the case studies to provide us with a basic level of confidence in the model. The simulation model appears to be in good agreement with the case studies in terms of electricity use. The simulation model for Seattle predicts electricity usage of 77 kwh/sf/yr, which is within 5% of the estimate for the Fred Hutchinson Center without the efficiency measures. The gas usage is more difficult to compare. The case studies report less than 2 kbtu/sf/yr of measured gas use with the efficiency measures. The case study for the laboratory in Albuquerque predicted total energy use at 428 kbtu/sf/yr without the efficiency measures. Estimating the electricity use at 75 kwh/sf/yr leaves 192 kbtu/sf/yr of gas use. Albuquerque s winters are more severe than Seattle s and less severe than Denver s. However, it is unclear from the case studies what fraction of gas usage is for loads other than space heating, and whether or not there are humidity controls. To understand the simulation model results, we compared the space heating energy use with a calculation of the energy needed to heat the air over the year using heating degree days (base 65). Because this is effectively a 24-hour facility, the heating degree days provide a good benchmark for estimating space heating loads from the ventilation requirements. Under winter conditions in the middle of the night, the balance point temperature is 65 F, assuming internal loads of 4 W/sf. The flow rate through the fans 13

16 has been determined under design conditions; it is 2 cfm/sf in the core zones and 2.7 cfm/sf to 3.2 cfm/sf (Table 1.5) in the perimeter zones. The space heating load is calculated using the following equation: Q = 1.8 FLOW ( T), (Equation 1.1) where Q is the heat loss, FLOW is the ventilation rate in cfm, and T is the temperature difference in F. The factor 1.8 represents the density and specific heat capacity of air at standard pressure and temperature multiplied by 6 to convert units. In Denver, standard air conditions are corrected for high altitude by dividing by To calculate the annual heating load, substitute heating degree days for the temperature difference in Equation 1.1. Space Heating (kbtu/sf/yr) Heating Loads Minneapolis* Denver Seattle Atlanta Space Ht Load () Vent Ht Load (HDD65) Space Ht Load (No Humidity) Figure 1.3 Space heating loads in Denver and Seattle. Figure 1.3 compares the space heating load predicted by DOE-2 with the ventilation heating load calculated using heating degree days. The space heating load from DOE- 2 is shown with and without humidity controls. The simulation results for the models without humidity control are within 5% of the space heating load calculated from heating degree days, base 65. Clearly, humidification of the air in Minneapolis and Denver has a significant influence on energy use. Another issue we investigated is the DOE-2.2 equipment part-load efficiency curve. Figure 1.4 shows the part-load efficiency curves used in DOE-2 for chillers and boilers. The energy-input ratio (EIR) is the inverse of the coefficient of performance (COP) and is the ratio of the energy input to the cooling output (unitless). The DOE-2.2 curve for the chiller is compared with actual data from an energy-efficiency chiller and with the cubic equation used to fit the actual data. The EIR has been normalized by the EIR at full-load (.5 kw/ton). Actual chiller performance can be characterized in version 41 of DOE-2.2 more precisely; this feature should be used in modeling actual buildings. The curve for the boiler is compared with a one-to-one curve that represents a boiler with no change in efficiency at part-load operation. The heating-input ratio (HIR) is the 14

17 ratio of the energy input to the heating output; it has been normalized in the curve with respect to the full-load efficiency (8%). For the boilers, part-load efficiency drops off significantly at a partial load of less than 3%. The simulation results show that the average annual efficiency for space heating in all the models is 68% in Atlanta, 71% in Seattle, and 72% in Denver and Minneapolis. From this, we concluded that the DOE- 2.2 curves are reasonable for this study. Chiller Part-Load Efficiency DOE-2.2 Boiler Part-Load Efficiency Normalized EIR Actual Cubic DOE2 Nomalized HIR DOE Part Load Part Load 1 Figure 1.4 Part-load efficiency curves for chillers and boilers. The models are predicting acceptable results. The comparison between the model heating loads and those calculated using heating degree days shows good agreement, assuming no humidity controls. The models predict much higher spacing heating energy use than the case studies do; this can be attributed to humidity controls, the high percentage of laboratory space in this building (7%), constant-volume fans, and design assumptions Annual Energy Use Although equipment loads dominate energy use, in general, energy end uses and monthly energy use vary with climate. A breakdown of electricity end uses in Seattle is shown in Figure 1.5. The equipment accounts for more than 5% of electricity use in all four climates. Lights average 5% to 6% of the electricity use, fans are in the 25% to 28% range, pumps are 2% to 4%, and space cooling varies from 4% in Seattle to 17% in Atlanta (Figure 1.6). Although it is important to characterize the equipment loads 15

18 Pumps 2% Case in Seattle Fans 28% Lights 6% accurately for sizing equipment, addressing the energy efficiency of this specialized equipment is difficult, and it does not necessarily apply from one building to the next. Although other end uses appear to account for a small percentage of total energy use, they are nevertheless energyintensive. Heat Reject % Space Cool 4% Equip 6% Figure 1.5 Electricity end uses in Seattle. Electricity (kwh/sf/yr) Electricity End Uses Figure 1.7 excludes the equipment load from electricity use to demonstrate the significance of space cooling and fan electricity use. Lights Equip Space Cool Heat Reject Pumps Fans In all four climates, annual fan Figure 1.6 Electricity end uses in all electricity use averages 21.7 kwh/sf. climates. Annual space cooling electricity use varies; it is 2.8 kwh/sf in Seattle, 5 kwh/sf in Denver, 8.2 kwh/sf in Minneapolis, and 15.6 kwh/sf in Atlanta. Electricity (kwh/sf/yr) Electricity End Use Without Equipm ent Load Lights Space Cool Heat Reject Pumps Fans Figure 1.7 Electricity end uses excluding equipment load. Another significant end use is space heating. The space heating energy includes energy used to heat the building and maintain specified humidity levels. For a benchmark, note that total energy use in office buildings is usually less than 1 kbtu/sf/yr. Space-heating energy use alone for this 24-hour laboratory with a constant ventilation rate is 4 to 8 times the energy use of an office building. Figure 1.8 presents space heating energy use with humidity controls (base case) and without humidity controls. The model for Denver, which has the driest climate, shows 16

19 the greatest humidification load. The results show that the minimum humidity levels in all four climates dip below 3% relative humidity (RH) for a significant number of hours. Humidity levels are in the 2%-29% RH range for 26 hours in Atlanta and 65 hours in Denver. We infer that the majority of these hours are above 28% RH. Annual Space Heating Energy Energy Use (kbtu/sf) Space Heat No Humidity Controls Figure 1.8 Annual space heating energy use with and without minimum humidity control Monthly Energy Use Laboratory buildings have such high internal loads from equipment that monthly electricity use varies by less than 15% over the year in milder climates like those of Seattle and Denver (Figure 1.9). However, electricity use is 5% higher in the summer, however, in hotter climates like Atlanta s. There is much greater variation in gas usage from month to month in all four climates (Figure 1.1). The heating load in the summer months is the result of subcooling and reheat for dehumidification. Monthly Electricity Use Electricity (kwh/sf/mo) Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Figure 1.9 Monthly electricity use in each of the climates. 17

20 Monthly Space Heating Energy Use Heating Energy (kbtu/sf/mo) Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Figure 1.1 Monthly space-heating energy use Peak Electricity Demand Peak electricity demand is two to three times greater than that generally found in commercial buildings (Figure 1.11). The gross equipment load itself is 7.2 W/sf, the fans contribute 2.5 W/sf, and lighting adds another 1.4 W/sf. The difference in peak demand between the climates is attributable to the difference in space cooling. 18

21 Monthly Peak Electricity Demand Peak Demand (W/sf) Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Figure 1.11 Monthly peak electricity demand for all climates. Another important consideration is on-peak and off-peak demand. For the electricity rates, the on-peak time is 8 a.m. to 1 p.m., Monday through Friday. Figure 1.12a, b, c, and d show the on-peak and off-peak electricity demand for each month. Peak Demand in Minneapolis Peak Demand in Denver Electricity (kw) Electricity (kw) J an Ma r Ma y Ju l S ep Nov J an M ar Ma y Ju l Sep Nov On-Peak Off-Peak On-Peak Off-Peak Figures 1.12a and b On-peak and off-peak electricity demand for Minneapolis and Denver. 19

22 Peak Dem and in Atlanta Peak Dem and in Atlanta Electricity Demand (W/sf) Electricity Demand (W/sf) Jan Mar May Jul Sep Nov Jan Mar May Jul Sep Nov On-Peak Off -Peak On-Peak Off -Peak Figures 1.12c and d On-peak and off-peak electricity demand for Seattle and Atlanta Supply Air The supply air is a minimum of 2 cfm/sf of outside air for the base case. The design cooling load in the perimeter zones requires 2.6 to 2.7 cfm/sf of supply air; 2 cfm/sf is sufficient in the core zones, except in Denver (Table 1.5). The ventilation requirements in Denver are higher because of its elevation. 4. Internal Loads and Supply Air Flow Supply Air Flow (cfm/sf) Internal Load (W/sf) Figure 1.13 Supply air flow at 55 F to meet internal loads. Figure 1.13 gives the supply air flow (cfm/sf) at 55 F required to offset internal loads. The internal loads in the base building average 1 W/sf during occupied periods, although the design loads are in the 15 W/sf range in the perimeter zones. 2

23 1.5.6 Sensible and Latent Loads The base case model includes humidity controls, so latent cooling and latent heating loads have a greater impact on energy use than they do in a building without such controls. The latent cooling load refers to the energy content of the moisture in the supply air that exceeds the maximum humidity requirement. The latent heating load is the energy content of the moisture needed to meet the minimum humidity requirement. DOE-2 reports the sensible heat ratio, which is the ratio of the sensible energy load to the total energy load (sensible plus latent). The lower the sensible heat ratio is, the more moisture there is in the air, and the higher the relative humidity is. In Denver, the sensible heat ratio is 1. (i.e., no latent cooling load) at the cooling peak in each month. In Seattle, the sensible heat ratio does not drop below.7 during the cooling peaks. In Minneapolis, the sensible heat ratio averages.42 at the cooling peak from July through September. In Atlanta, the sensible heat ratio drops as low as.38 in July; it is around.5 during other summer months. To dehumidify the air in Minneapolis and Atlanta, the supply air is subcooled to remove moisture from the air and reheated to bring the air temperature back up to the minimum 55 F supply air temperature. In dry months, heating energy is required to evaporate moisture into the air to meet minimum humidity requirements, i.e., latent heating energy. Atlanta requires latent heating energy from October through March, whereas Minnesota and Seattle require it from October through May. Denver has year-round latent heating requirements, because its climate is so dry. Figure 1.8 compares space heating requirements with and without humidity controls. To add moisture to the air, the hot water loop serves a pan heat exchanger through which moisture is evaporated into the air Energy Costs A rule of thumb for commercial buildings is that energy costs average $1/sf/yr. For laboratories, however, the cost is $5 to $1/sf/yr. The simulation models reflect this, with electricity costs averaging $4/sf/yr. Gas costs range from $2/sf/yr in Seattle to $5/sf/yr in Minneapolis (Figure 1.14). d on the assumed utility rates, the cost for electricity averages $.5/kWh; that for gas is $.6/therm. The demand charges for electricity are 59% of the total electricity charges. The structure of the utility rates has a big impact on energy costs and varies from utility to utility. 21

24 Annual Energy Costs $/sf/yr $1 $9 $8 $7 $6 $5 $4 $3 $2 $1 $ Gas Elec (kwh) Elec Demand Figure 1.14 Annual electricity and gas costs for all climates. 22

25 CHAPTER 2. ENERGY EFFICIENCY STRATEGIES Table 2.1 presents the energy efficiency strategies we considered. These strategies focus on reductions in fan energy use, energy recovery opportunities, and evaporative cooling. The effect of tighter humidity controls with and without an enthalpy wheel is also evaluated. In addition, we investigated the impact of equipment power density assumptions (plug loads) on mechanical system sizing. Table 2.1 Energy Efficiency Strategies Measure Run 1 Run 2 Minimum setting: 2 cfm/sf occupied, 1 cfm/sf unoccupied (24 Ventilation Rates hrs/day) setting Static Pressure Drop Energy Recovery: Enthalpy Wheel Heat Pipes Run-Around Loop Chiller Energy Recovery Evaporative Cooling Minimum setting: 2 cfm/sf (24 hrs/day) 5.2 in. supply, 2 in. exhaust None None None None None 4 in. supply, 1.5 in. exhaust at 2 cfm/sf.75 sensible effectiveness;.75 latent effectiveness at.8 in. w.g. pressure drop.48 effectiveness at 1 in. pressure drop.6 effectiveness at 1 in. pressure drop.8 fraction of heat recovered from condenser water Direct evap. w/.8 effectiveness and.1 in. w.g. pressure drop Variable air volume with 1 cfm/sf minimum 3 in. supply, 1 in. exhaust at 2 cfm/sf Same as previous with system Water-side economizer with.8 effectiveness 4% RH Min/ 5% RH Max w/ Enthalpy Wheel 4% RH Min/ 5% RH Humidification 3% RH Min/ 6% RH Max Max (Also ran 2% RH Min/ 6% RH Max) Plug Loads in Lab Space 12 W/sf 8 W/sf 4 W/sf Table 2.2 reiterates energy use statistics for the base case building. The energy efficiency measures are compared with the base case for each climate. We conclude by looking at the most efficient strategies for each climate. Peak Demand (W/sf) Table 2.2 Case Building Energy Use Electricity Use (kwh/sf/yr) Total Energy Use (kbtu/sf/yr) Annual Electricity Cost ($/sf) Annual Gas Cost ($/sf) Annual Energy Cost ($/sf) Gas Use (kbtu/sf/yr) Minneapolis $4.3 $5.2 $9.5 Denver $4. $4.8 $8.8 Seattle $3.9 $2.6 $6.6 Atlanta $4.7 $2.2 $6.9 23

26 2.1 Ventilation Rates The base case building has a constant-volume air system. The design flow rates are shown in Table 2.3 for the core and perimeter areas. Under peak cooling conditions (not design conditions), the cooling load is 1 W/sf and requires 1.6 cfm/sf of 55 F air to cool. During unoccupied periods, the cooling load is less than 6 W/sf and requires only 1 cfm/sf of 55 F air to cool (Figure 1.13). The difference between the design condition of more than 2 cfm/sf and a minimum of 1 cfm/sf provides opportunities to reduce energy use for fans, space cooling, and space heating. Table 2.3 Design Supply and Exhaust Flow Rates Core Perimeter (cfm/sf) (cfm/sf) Minneapolis Denver Seattle Atlanta A simple control strategy is to reduce the minimum supply flow from 2 cfm/sf to 1 cfm/sf during unoccupied periods (CFM21). A more efficient approach is a system with reheat to reduce the supply air flow in response to varying loads and varying ventilation requirements, as would occur with variable-volume fume hoods. The system is modeled with a minimum outdoor air setting of 1 cfm/sf and an increase in static pressure drop of.5 in. w.g. to account for the losses associated with the variablespeed drive. DOE-2.2 calculates the load in each zone and determines the necessary supply air flow. The fans have variable-speed drives, and DOE-2 calculates the fan energy with respect to the lower flow rate and lower static pressure drop. Electricity Use (kwh/sf/yr) CMF21 Electricity End Uses CMF21 CMF21 CMF21 Lights Space Cool Heat Reject Pumps Fans Figure 2.1 shows the electricity end uses for the base case, the flow setback case (CFM21), and the case, excluding equipment plug loads. Flow setback reduces the annual electricity use by 3 kwh/sf in Minneapolis, 2 kwh/sf in Denver, 1 kwh/sf in Seattle, and 5 kwh/sf in Atlanta. The system reduces annual electricity use by 8 kwh/sf in Minneapolis, 5 kwh/sf in Denver, 6 kwh/sf in Seattle, and 12 kwh/sf in Atlanta. Both strategies reduce electricity use for Figure 2.1 Electricity end uses for the fans. Figure 2.2a shows fan electricity three cases. use, and Figure 2.2b shows savings in comparison to the base case fan electricity use, 21.7 kwh/sf. The supply air flow setback strategy reduces the fan electricity use by 3 kwh/sf/yr in Minneapolis, 2 kwh/sf/yr in Denver and Seattle, and 24

27 4 kwh/sf/yr in Atlanta. The system reduces fan electricity use by 7 kwh/sf/yr in Minneapolis, 5 kwh/sf/yr in Denver, 7 kwh/sf/yr in Seattle, and 9 kwh/sf/yr in Atlanta. Electricity Use (kwh/sf/yr) Fan Ele ctricity Us e CFM21 Electricity Savings (kwh/sf/yr) Fa n Ele c t r ic it y S avin g s CF M 2 1 VA V Minn eapolis D e nv er S e a ttle A tlanta Figure 2.2a and b Annual fan electricity use (a) and savings (b) resulting from reducing supply air flow with setback controls (CFM21) and. Electricity Use (kwh/sf/yr) Space Cooling Electricity Use CFM21 The effect of flow setback on annual space cooling savings is negligible (Figure 2.3). The most significant cooling savings occur in Atlanta, where the system reduces cooling electricity use by 3.1 kwh/sf. This is a reduction of 2% in cooling electricity use, although it is just one-fourth of the savings resulting from reducing fan energy. In Seattle, space cooling is adversely affected as a result of running the chillers at lower part loads. Figure 2.3 Annual electricity use for space cooling resulting from reducing supply air flow with setback controls (CFM21) and (). Peak electricity demand does not change with the supply air flow setback. It drops by 1 W/sf with in Minneapolis and Denver and by 2 W/sf in Atlanta. In Seattle, the peak reduction is.5 W/sf. Figures 2.4a and b show the space heating energy use and savings for each climate. The supply air flow setback reduces space heating energy use by 2% in Minneapolis and Denver, 1% in Seattle, and 6% in Atlanta. The system reduces space heating energy use by 19% in Minneapolis, 11% in Denver, 23% in Seattle, and 28% in Atlanta. 25

28 Space Heating Energy Use Space Heating Energy Savings Energy Use (kbtu/sf) CFM21 Energy Savings (kbtu/sf) CFM21 Figure 2.4a and b Space heating energy use and annual energy savings resulting from reducing space heating energy with flow reduction strategies. Energy Use (kbtu/sf/yr) CFM21 Total Energy Use CFM21 CFM21 Elec Gas CFM21 Figure 2.5 Total energy use for the base case and flow reduction strategies. In terms of total energy use, the electricity energy use is comparable to the gas energy use in Atlanta, whereas in Minneapolis, gas energy use is nearly 4 times the electricity energy use (Figure 2.5). However, the cost per Btu of electricity is nearly 3 times the cost per Btu of gas in this analysis, which translates into 3 times the savings for every Btu of electricity conserved. The impact of the flow reduction strategies on total energy use is not as great as expected. The model assumes high internal gains from plug loads, which limits the opportunity to reduce air flow. The base energy costs range between $6.5/sf in Seattle to $9.5/sf in Minneapolis. With the flow setback controls, the electricity savings are $.2/sf in Minneapolis, $.1/sf in Denver, $.1/sf in Seattle and $.2/sf in Atlanta. The electricity cost savings from the system are $.4/sf in Minneapolis, $.3/sf in Denver and Seattle, and $.6/sf in Atlanta (Figure 2.6a and b). 26

29 The gas savings from the flow setback are $.1/sf in Minneapolis, $.1/sf in Denver, $.3/sf in Seattle, and $.1/sf in Atlanta. The system saves $1/sf in Minneapolis, $.5/sf in Denver, $.6/sf in Seattle, and $.6/sf in Atlanta (Figure 2.6a and b). Annual Energy Cost Annual Energy Cost Savings Energy Cost ($/sf/yr) $12. $1. $8. $6. $4. $2. $. CFM21 CFM21 Electricity CFM21 Gas CFM21 Cost Savings ($/sf/yr) $1.6 $1.4 $1.2 $1. $.8 $.6 $.4 $.2 $. CFM21 CFM21 CFM21 Gas Electricity CFM21 Figure 2.6a and b Annual energy costs and savings for the base case and flow reduction strategies. The system saves $.2/sf to $.3/sf in electricity costs and $.4/sf to $.9/sf in gas costs over the flow setback controls. On average, cost savings from the flow setback are 3% of the base case costs, and savings from the system are 14% of the base case costs. Boiler Horsepower (hp) Reduction in Boiler HP with Another advantage of the system is a potential opportunity to reduce the size of the heating system. The heating system would be designed to meet the heating load using a ventilation requirement of 2 cfm/sf, rather than the design flow determined under cooling design conditions. The potential reduction in boiler size varies from 1 to 3 hp in these four climates (Figure 2.7) and results in cost savings of $25, to $5, at $25/hp. Figure 2.7 Potential reduction in boiler size with system. 27

30 2.2 Fan Static Pressure Drop Fan energy use is calculated from the following equation: Energy = Static Pressure Drop *Flow *.746 / η / 6354, (Equation 2.1) Fan Energy (W/cfm) Fan Energy (76%Efficiency) Total Static Pressure Drop (in. w.g.) Figure 2.8 Fan energy with respect to static pressure. where Static Pressure Drop is the pressure drop associated with coils, filters, and ducts; Flow is the air flow rate in cfm; the factor.746 converts horsepower to watts; η is the combined efficiency of the motor and fan; and the factor 6354 converts units to horsepower. The result is energy use (Energy) in W/cfm. For the base case, the fans are constant-volume and have a total efficiency of 76%. The motor efficiency is 95%. The supply fans see a total static pressure drop of 5.2 in. w.g. and the exhaust fans see a total static pressure drop of 2 in. w.g. Figure 2.8 shows the fan energy use in W/cfm with respect to static pressure drop. Reducing the static pressure drop reduces fan energy use. The static pressure drop can be reduced through the design and use of larger ducts and coils and filters with a lower pressure drop. Electricity End Uses Space Heating Energy Use Electricity Use (kwh/sf/yr) SP4 SP3 SP4 SP3 SP4 SP3 SP4 SP3 MINNEAPOLIS' Denver Seattle Atlanta Space Cool Heat Reject Pumps Fans Energy Savings (kbtu/sf) SP4 SP3 Figure 2.9 Electricity end uses for the base case, supply static pressure of 4 in. w.g. and exhaust static pressure of 1.5 in. w.g (SP4), and supply static pressure of 3 in. w.g. and exhaust static pressure of 1. in. w.g. (SP3). Figure 2.1 Space heating energy use for the base case, supply static pressure of 4 in. w.g. and exhaust static pressure of 1.5 in. w.g (SP4); supply static pressure of 3 in. w.g. and exhaust static pressure of 1 in. w.g. (SP3). 28

31 To demonstrate the potential electricity savings resulting from reducing the static pressure drop, we simulated two cases: (1) supply static pressure drop of 4 in. w.g. and exhaust of 1.5 in. w.g., and (2) supply static pressure drop of 3 in. w.g. and exhaust of 1. in. w.g. The input to DOE-2.2 is the design static pressure drop, fan and motor efficiency, and the fan curve. The reduction to 4 in. w.g. and 1.5 in. w.g. saves 5 kwh/sf/yr, and the reduction to 3 in. w.g. and 1 in. w.g. saves 1 kwh/sf/yr (Figure 2.9). There is an increase in space heating energy use because less fan energy is being added to the air stream (Figure 2.1). The peak electricity demand is reduced by.7 W/sf in the first case and by 1.3 W/sf in the second case. The fan system is constantvolume, so the fan energy use is constant for the base case. Energy Cost ($/sf-yr) Annual EnergyCost $12. Electricity Gas $1. $8. $6. $4. $2. $. SP4 SP3 SP4 SP3 SP4 SP3 SP4 SP3 Cost Savings ($/sf/yr) $.5 $.4 $.3 $.2 $.1 $. -$.1 -$.2 Annual Energy Cost Savings Gas Electricity SP4 SP3 SP4 SP3 SP4 SP3 SP4 SP3 Figure 2.11a and b Annual energy cost and cost savings from reducing supply static pressure from 5.2 in. w.g. (base) to 4 in. w.g. (SP4) and 3 in. w.g. (SP3). The annual net cost savings are $.17/sf in the first case and $.32/sf in the second case (Figure 2.11a and b). The increase in gas costs is $.8/sf with the 4 in. w.g. static pressure (SP4) and $.13/sf with the 3 in. static pressure (SP3). With a system, the flow rate and static pressure drop are reduced. As shown in the previous section, the result is lower electricity and gas usage. The annual cost savings with a system are $1.4/sf in Minneapolis, $.8/sf in Denver, $.9/sf in Seattle, and $1.2/sf in Atlanta. 29

32 2.3 Energy Recovery Figure 2.12 Example of run-around loop energy recovery system. Energy recovery is often considered for laboratories because of the high outside air ventilation rates. There are many options available for air-to-air energy recovery; they are summarized in the ASHRAE HVAC Systems and Equipment Handbook (2). for this analysis, we considered enthalpy wheels, which recover sensible and latent energy, and heat pipes and run-around loops (Figure 2.12), which recover sensible energy only. Energy recovery from the chiller to the hot water loop is also included. For the performance of the enthalpy wheels and heat pipes, we referred to the ARI Airto-Air Recovery Ventilation Equipment Certified Products Directory (March 22) and Des Champs and Semco technical representatives. For the run-around loop, we used the ASHRAE HVAC Systems and Equipment Handbook (2). The effectiveness of the air-to-air recovery devices is defined as the ratio of the actual energy recovered to the theoretical energy that could be recovered. Sensible energy is the energy associated with a temperature difference. The sensible effectiveness is proportional to the ratio of the difference between the dry bulb temperature of the outside air and supply air to the difference between the dry bulb temperature of the exhaust air and the outside air. Latent energy is the energy of the moisture, and in this case the moisture in the air. The latent effectiveness is proportional to the ratio of the difference between the humidity ratio of the outside air and the supply air to the difference between the humidity ratio of the exhaust air and the outside air. The DOE-2.2 model of energy recovery ventilators (ERV) has recently been added and is still being tested. DOE-2.2 (version h) is zeroing out the humidity ratio for the air leaving sensible heat recovery devices, but the model predicts it correctly for devices with sensible and latent heat recovery. In order to model heat pipes and run-around loops with humidity controls, we used the enthalpy-hx option in DOE-2.2 and set the latent effectiveness to.5. 3